No. 25-3582

Appeal from the United States District Court for the Northern District of Ohio at Cleveland. No. 1:16-cv-00917 — Jennifer Dowdell Armstrong, Magistrate Judge.

Argued: April 30, 2026

Decided and Filed: May 27, 2026

Before: SUTTON, Chief Judge; CLAY and MURPHY, Circuit Judges.

ARGUED: James R. Saywell, JONES DAY, Cleveland, Ohio, for Appellant. Douglas C. Rennie, UNITED STATES DEPARTMENT OF JUSTICE, Washington, D.C., for Appellee. ON BRIEF: James R. Saywell, Elizabeth A. Dengler, Andrew S. Rumschlag, JONES DAY, Cleveland, Ohio, Laura Kingsley Hong, Brendan Kelley, TUCKER ELLIS LLP, Cleveland, Ohio, for Appellant. Douglas C. Rennie, Michael J. Haungs, UNITED STATES DEPARTMENT OF JUSTICE, Washington, D.C., for Appellee.

SUTTON, C.J., delivered the opinion of the court in which CLAY and MURPHY, JJ., concurred. MURPHY, J. (pp. 20 – 30), delivered a separate concurring opinion, in which SUTTON, C.J., joined.

Page 2SUTTON, Chief Judge. The Internal Revenue Service, it’s often said, requires taxpayers to “turn square corners.” Rock Island, Ark. & La. R.R. Co. v. United States , 254 U.S. 141, 143 (1920). Complex statutes, book-length regulations, and too-many-part tests are an unfortunate reality of those turns. Yet the Internal Revenue Code, for all of the challenges of identifying taxable and nontaxable events before people act, still strives to make our nation’s tax system 'accessible to everyone with the time and patience to pore over its provisions.' Summa Holdings, Inc. v. Comm'r , 848 F.3d 779, 781 (6th Cir. 2017). Congress and the citizens it represents prefer seen corners to unseen ones.

In today's case, the government seeks to impose a $39 million judgment, including interest and penalties, on Flight Options, a fractional-share jet company, for failing to collect a tax on fixed fees it charged to pay for the overhead and management of its clients’ private jets. The Internal Revenue Code imposes a 7.5% excise tax on the 'amount paid for' domestic “transportation by air,” 26 U.S.C. §§ 4261(a), 4262(a)(1), what the statute called a “ticket tax” at all relevant times of this dispute, id. § 4261(e)(1)(C), (e)(5) (2012). The Code imposes the tax on the ticket buyer. But the Code makes the ticket seller liable for any tax it fails to collect on behalf of the government. Flight Options determined that the tax applies only to usage charges for each flight a client takes, not to fixed fees it charges its clients for overhead and management of its fractional jet business.

The district court disagreed. It held that Flight Options should have collected taxes on the fixed fee charges as well and that, having failed to collect them, must pay the balance, with interest and penalties to boot. Because the ticket tax applies only to usage charges for each flight and not fixed charges for overhead and management costs, we reverse.

Page 3Flight Options offers individuals a way to fly by private jet without having to buy a plane themselves. It provides two models.

The first is a fractional-share jet ownership program, which sells clients a percentage interest in an aircraft. Through the program, a client buys access to the aircraft in proportion to his ownership stake. In addition to the initial purchase price for the ownership share, Flight Options charges its fractional jet owners a fixed fee that applies whether they fly or not. Fractional jet owners pay this monthly management fee in proportion to their aircraft ownership. The fee covers the costs of owning an airplane, such as leasing the hangar, conducting inspections and repairs, and securing insurance. It also covers operational overhead expenses, which include pilot and crew salaries, FAA paperwork, and Flight Options’ administration of the fractional jet owner program.

The second service that Flight Options offers is called the Jet Club Membership program. It charges clients an upfront membership fee in exchange for giving them the option to purchase flight time on private planes owned by Flight Options — not unlike a charter flight service.

Flight Options also charges fractional jet owners and jet club members usage fees for any time spent flying. The payments go to costs incurred for specific flights, such as fuel, flight planning, and weather and communication services. Flight Options also bills its clients for the movement of an aircraft to meet them and for the time spent waiting if the client arrives late. These fees depend on whether the client uses the plane. If no flight is booked, none of these usage fees is charged.

The fractional-share jet industry began in the 1980s with Executive Jet, now known as NetJets. At the outset, the general practice of the companies was not to collect this excise tax on either usage fees or fixed overhead and management fees. The industry saw itself as selling ownership interests, not transportation tickets. Exec. Jet Aviation, Inc. v. United States , 125 F.3d 1463, 1465 – 68 (Fed. Cir. 1997). That changed in 1992, when the IRS told Executive Jet to begin collecting the excise tax on usage fees. Rather than issue a regulation or revenue ruling to that effect, the IRS issued a technical advice memorandum. See IRS Tech. Adv. Mem. 93 – 14 – 002Page 4 (Dec. 22, 1992). A technical advice memorandum binds the IRS with respect to the receiving taxpayer and, by statute and regulation, does not bind other taxpayers or the IRS itself in other cases. See 26 U.S.C. § 6110(k)(3); 26 C.F.R. § 601.601(d)(1).

Executive Jet disagreed with the memorandum and went to court. It lost. In 1997, the Federal Circuit ruled that the excise tax applied to usage fees and that Executive Jet had to collect taxes on them. Exec. Jet , 125 F.3d at 1469. As the Federal Circuit acknowledged, however, the IRS had never applied the tax to charges for fixed overhead costs or general management expenses. Id. at 1467.

Enter Flight Options. It began offering fractional-share jet ownership, jet club membership, and related services in 1998. By then, the industry had begun to collect excise taxes on usage fees but continued not to collect the taxes on fixed fees for management and overhead.

In 2004, the IRS changed course. It determined that the private jet industry had to collect the tax on fixed management and overhead fees as well. The IRS did not enact a regulation or issue a revenue ruling to this effect. It instead announced this position through another technical advice memo, this time addressed to Bombardier, a competing fractional-share jet provider. See IRS Tech. Adv. Mem. 2004 – 42 – 5048 (Feb 17, 2004); Bombardier Aerospace Corp. v. United States , 831 F.3d 268, 280 (5th Cir. 2016). By 2008, the IRS adopted a policy of enforcing the tax on all fixed fees for management and overhead and, to that end, began audits of Flight Options and other fractional-share jet operators. See Bombardier , 831 F.3d at 280.

The industry resisted this change in taxation policy. After a series of unsuccessful meetings with IRS agents asking for clarity under the new program, four industry members sued the government for abatement of the excise tax assessed on fixed-fee payments. The group included Flight Options, which challenged the IRS’s assessment of taxes for the periods between March 2004 and December 2007. See Flight Options LLC v. United States , No. 11-cv-01531 (N.D. Ohio filed July 25, 2011).

The industry record in this litigation was all over the map. NetJets secured a total abatement of the collection for fixed fees. See NetJets Large Aircraft Inc. v. United States ,Page 5 2015 WL 7784925, at *11 (S.D. Ohio Nov. 12, 2015). Bombardier had to pay for two years of uncollected fixed fees, given the reality that the IRS directed the 2004 technical advisory memorandum to Bombardier. See Bombardier , 831 F.3d at 280 – 81, 284. Another industry provider settled with the IRS, with the IRS conceding that the provider did not have to collect any fixed fees. The IRS also settled with Flight Options, conceding that it did not have to pay 80% of the tax on the fixed fees for the years 2004 through 2007. Throughout the litigation, Flight Options and the rest of the industry continued to collect taxes solely on usage fees.

In 2012, the controversy came to a partial end when Congress amended the Tax Code to exempt fractional-share jet owners from all taxes under § 4261, whether applied to usage fees or fixed fees. See Pub. L. No. 112-95, § 1103, 126 Stat. 11, 151 (2012). The litigation continued, however, over potential liability for the operators for failing to withhold the tax on fixed fees charged before the 2012 effective date.

That brings us to the present case. At stake is an IRS assessment of $24 million in uncollected taxes on fixed fees, plus interest and penalties, for a total of about $39 million against Flight Options during the period from January 1, 2009, through March 31, 2012. Flight Options sued the government to abate the assessment. The parties consented to have a magistrate judge decide the case. Relying on IRS revenue rulings, the magistrate judge decided that the fixed fees paid to Flight Options were “amounts paid for” “transportation by air.” R.90 at 27 – 34. The judge also imposed failure-to-collect penalties on Flight Options, concluding the company should have known the IRS would attempt to collect taxes on the fixed fees. Flight Options appealed.

This excise tax, enacted in 1956 and applicable to fractional-share jet operators until 2012, does not extend to the fixed overhead and management fees that Flight Options charged its owners during the relevant time period, as opposed to the usage fees that the company charged for each flight. All indicators of meaning — the text of the excise tax, the context in which it appears, the regulations, and the relevant canons of interpretation — lead to the same conclusion: The tax does not apply.

Page 6Text . Start with the language of § 4261. At the pertinent times, the statute called this excise provision a “ticket tax.” 26 U.S.C. § 4261(e)(1)(C), (e)(5) (2012); see Pub. L. No. 10534, § 1031, 111 Stat. 788, 931 – 32 (1997). The statutory label —a “ticket tax”— suggests that the excise tax applies at most to flight-by-flight usage charges -the kind of 'tickets' issued for flights when Congress passed the statute — not fixed overhead charges or management fees for fractional-share jet owners.

The operational language of the tax takes the reader in the same direction. The statute covers the “amount paid for taxable transportation of any person.” 26 U.S.C. § 4261(a). It then defines 'taxable transportation' as 'transportation by air.' Id. § 4262(a)(1). This covered service — taking someone from point A to point B by air — speaks to a flight-by-flight tax, not a tax for the overhead of the business that provides the service.

While the statute does not define 'transportation,' contemporary dictionaries confirm what common understanding would suggest. The term refers to the act of moving persons from one place to another. See, e.g. , Webster’s Third New International Dictionary 2430 (1976) (“an act, process, or instance of transporting”); Ballentine’s Law Dictionary 1294 (3d ed. 1969) (“The carriage of persons or property from one point to another”); Black’s Law Dictionary 1670 (4th rev. ed. 1968) (“The removal of goods or persons from one place to another”). The word “for,” then and now, means “to obtain.” American Heritage Dictionary 512 (1969); see also Black’s Law Dictionary, supra , 772 ('in exchange for'). Taken together, 'payment for transportation' naturally means '[t]he fare paid by a passenger on train, bus, [or] plane,' Ballentine's Law Dictionary, supra , 926, what the statute itself calls a “ticket tax.” “Fares” paid for the “instance” of movement do not naturally describe fixed monthly or one-time membership fees for overhead and management.

Context . Statutory context reinforces this understanding. The statute applies the tax to specific 'domestic segments' of 'taxable transportation.' 26 U.S.C. § 4261(a), (e)(1)(C)(ii) (2012). It talks of where (in the United States) and when (after a certain date and before 2012) the transportation must 'begin[].' Id. §§ 4261(j)(1), 4262(a)(1) (2012). Congress's 1970 amendment to the statute maintains this meaning by defining transportation to include layovers and deadhead service — flying an empty plane to meet crew or passengers. See id. § 4262(d); cf.

Page 7Bhd. of Locomotive Eng’rs v. Atchison, Topeka & Santa Fe R.R. Co. , 516 U.S. 152, 154 (1996). Both layovers and deadhead flights occur during or in preparation for specific flights. To the extent that Congress has ever taxed fees in “connection with transportation,” it imposed that tax on charges for seating and sleeping accommodations — expenses incurred on a flight-by-flight basis. 26 U.S.C. § 4261(c) (1962).

The exclusions echo this flight-by-flight trigger for taxation. From the outset, the ticket tax has excluded two sets of payments “for transportation” from the tax. The first is payments for foreign transportation, say a flight from Columbus to London. Id. § 4262(b), (c). The second set is payments for the domestic transportation of non-persons, such as moving a pet from Cincinnati to Seattle. Id. § 4261(a); see 26 C.F.R. § 49.4261-8(f) (2012). Each excluded charge operates on the same axis as a usage charge, a domestic flight from one point to another for a person.

The statute contains one other relevant feature supporting the inference that the ticket tax applies on a flight-byflight basis. Section 4261 is a “collect[ed] tax,” meaning that it is paid by the passenger purchasing the ticket but collected by the carrier providing the transportation. 26 U.S.C. § 4291. The carrier must then remit the collected proceeds to the IRS twice a month. See Kaucky v. Sw. Airlines Co. , 109 F.3d 349, 350 (7th Cir. 1997); 26 C.F.R. § 40.6302(c)1(a)(1). That all points to a form of tax collection for payments corresponding to a particular use. The flier pays for a specific flight with a specific ticket, and the statute presumes the collector will immediately know whether that ticket incurs the tax. This system would swiftly grind to a halt if the tax collector didn't know, or couldn't know, what aspects of the management or overhead fees would count.

An adjacent tax adds a similar hint. Section 4251 imposes a telephone tax that reaches all “amount[s] paid for communication services.” 26 U.S.C. § 4251(a)(1). Section 4261 likewise could have taxed payments for transportation services . But Congress chose instead to use the term “transportation” alone and not to extend it to transportation services. If “Congress knows how to say something but chooses not to, its silence' deserves respect. Black Farmers & Agriculturalists Ass’n v. Rollins , 154 F.4th 473, 477 – 78 (6th Cir. 2025) (quotation omitted).

Page 8The use of the narrower term suggests that Congress anticipated that the tax applies on a flightby-flight basis.

The regulations . The ticket-tax regulations, none of which is challenged as beyond the Secretary's authority, reinforce this conclusion. See 26 U.S.C § 7805(a) (delegating to the Treasury Secretary authority to “prescribe all needful rules and regulations”). The regulations say that the tax on 'transportation' does not apply to certain items if 'separable from the payment for the transportation of a person.” 26 C.F.R. § 49.4261-8(f) (2012). These include “[c]harges in connection with the charter of' 'an air conveyance,' such as 'parking,' '[de]icing,' 'sanitation,' 'dockage,' 'wharfage,' and other 'similar charges.' Id. § 49.42618(f)(5) (2012). The reality that the ticket tax, in the IRS’s view, does not apply to parking, de -icing, dockage and other necessary features of a single flight confirms that the tax does not cover what the IRS claims it does —every “reasonably necessary,” Appellee’s Br. 30, or other but -for cost of a flight, much less the management fees of running a fractional jet ownership company. As originally implemented in 1962, this regulation also excluded charges for ''layover' or ‘waiting time,’ [and] movement of equipment in deadhead service.” 26 C.F.R . § 49.4261-8(f)(5) (1962). When Congress overruled the “layover,” “waiting time,” and “deadhead” exceptions in 1970, Pub. L. No. 91-258, § 203(b), 84 Stat. 219, 239; see 26 U.S.C. § 4262(d), it did not overrule these other exceptions.

Other regulations confirm the statute’s focus on (at most) flight -by-flight usage fees. The IRS’s examples of “payments for transportation” include “ticket[s]” and other payments for “the right to transportation.” 26 C.F.R. § 49.4261-7(a); see id. § 49.42617(j). Its list of “additional charges' subject to the tax relate to flight -specific services such as 'changing the class of accommodations,' selecting a new 'destination or route,' and 'extending the time limit of a ticket.” Id. § 49.4261-7(c). Although the IRS classifies payments for the charter of an aircraft as an example of 'taxable transportation,' these include only the 'charter charges' for specific domestic flight segments beginning and ending in the United States, namely the kinds of usage fees that Flight Options already collects. Id. §§ 49.4261-7(h)(2); 49.4262-1(a), (c)(5) (2012). Consistent with this interpretation, the regulations likewise exclude payments 'in connection with' the charter of an aircraft for transportation, including those necessary for the general Page 9maintenance of the plane (“parking”), its ongoing ability to fly (“[de]icing,” “sanitation”), and any “similar charges.” Id. § 49.4261-8(f)(5) (2012).

In the light cast by these provisions, Flight Options’ fixed fees do not constitute payments for 'transportation by air.' The monthly and membership fees provide neither a 'fare' nor a 'ticket' nor 'the right to transportation.' They instead provide the option to purchase flight hours — an option that fractional owners and jet members may exercise only after paying the hourly usage fee. Unlike usage charges, which correspond to a specific flight, membership and monthly fees accrue “regardless of [users’] flying activity.” R.75 -7 at 22. Many of the services covered by the fixed fees, such as insurance, depreciation repairs, and ground inspections, correspond with the costs of owning a plane rather than flying one and, appropriately enough, scale in “direct[] proportion[]” to a fractional jet owner’s share of the fleet. Id . Still other costs covered by the fixed fees go toward the kinds of charges expressly excluded by the regulations: aircraft storage, hangar space, safety inspections, and interior maintenance.

The relevant taxpayer canons . Two taxpayer canons cement this conclusion. When it comes to tax laws, we have followed the “established rule” against “extend[ing] their provisions” 'beyond the clear import of the language' and 'enlarg[ing] their operations so as to embrace matters not specifically pointed out.” Gould v. Gould , 245 U.S. 151, 153 (1917); accord United States v. Wigglesworth , 28 F. Cas. 595, 596 – 97 (C.C.D. Mass. 1842) (Story, J.); see Am. Net & Twine Co. v. Worthington , 141 U.S. 468, 474 (1891) (requiring 'clear and unambiguous language”). “When the tax gatherer puts his finger on the citizen, he must also put his finger on the law permitting it.' OfficeMax, Inc. v. United States , 428 F.3d 583, 594 (6th Cir. 2005) (quoting Leavell v. Blades , 141 S.W. 893, 894 (Mo. 1911)); see Antonin Scalia & Bryan A. Garner, Reading Law 237, 299 – 300 (2012).

Related to this general canon is a specific one that comes into play when the IRS enlists a third party, such as Flight Options, to collect withholding taxes. In that setting, if neither the Tax Code nor binding IRS regulations provide 'precise and not speculative' notice of a tax withholding obligation, the IRS has no right to impose that obligation after the transactions have taken place. Central Illinois Pub. Serv. Co. v. United States , 435 U.S. 21, 31 – 32 (1978). The notice requirement stems from 'a matter of obvious concern.' Id. Flight Options faces stiff Page 10consequences for over-collecting and undercollecting the 'transportation' tax. If it fails to collect the tax “at the time payment for transportation is made,” it becomes secondarily liable for the tax. 26 U.S.C. § 4263(c). And its officers become personally liable for the tax under § 6672(a). Id. §§ 4261(d), 4263(d), 4291; see Nakano v. United States , 742 F.3d 1208, 1211 (9th Cir. 2014). Over-collection comes with risks, too. If uncorrected, the overcollection risks civil penalties from the government, see Sigmon v. Sw. Airlines Co. , 110 F.3d 1200, 1206 (5th Cir. 1997), and the specter of lawsuits from overpaying customers, see In re Air Transp. Excise Tax Litig. , 37 F. Supp. 2d 1133, 1136 – 38 (D. Minn. 1999). The Tax Code also imposes hurdles on correction, requiring the carrier to repay all of its customers (and carry the full tax) before it may file for a refund from the IRS. 26 U.S.C. § 6415(a).

Third-party collectors thus find themselves seated on a two-headed horse, at once eager to run from civil penalties for overcollection and to bolt from personal liability for undercollection. Commandeered to perform this tax-collection service on behalf of the government, third-party collectors need — indeed deserve — clear instructions. See Central Illinois , 435 U.S. 31 – 32; see also W. Rsrv. Acad. v. United States , 801 F.2d 250, 251 (6th Cir. 1986) (per curiam), adopting 619 F. Supp. 394, 400 – 01 (N.D. Ohio 1985). As a source for such instructions, they may look to the statutes, binding IRS regulations, and judicial authority to determine their collection responsibility. See Central Illinois , 435 U.S. at 31 – 32; see also Bombardier , 831 F.3d at 279 (looking to whether any 'regulation or ruling require[es] withholding”); HB & R, Inc. v. United States , 229 F.3d 688, 691 – 92 (8th Cir. 2000) (looking to whether regulations “expressly distinguish” amounts subject to tax); Am. Airlines, Inc. v. United States , 204 F.3d 1103, 1111 (Fed. Cir. 2000) (citing Central Illinois for the proposition that courts should look to “relevant statutes, regulations, and IRS pronouncements”). If these sources fall short of creating a “precise and not speculative” obligation, we do not require the third -party tax collector to “fill the gap” or, worse, allow the gaps to be filled “by judicial determination” after the flights have left the ground. Central Illinois , 435 U.S. at 31 – 33.

The only plain feature of this statute is that Flight Options lacked adequate notice of its tax-collection responsibility. Because the statute never clarifies how it would apply beyond the flight-by-flight context of a ticket tax, it never describes what types of fixed charges could count Page 11as paying “for transportation.” More, even if there were some discernible way to place this wide range of charges on one side or the other of the taxation line, Flight Options had no guidance for allocating them between taxable domestic transportation and nontaxable international transportation.

Today's dispute illustrates the import of this clarity obligation. In one corner, Flight Options has accepted that the excise tax applies to usage fees — the charges for each flight, what looks like the “ticket tax” the statute covers. In the other corner, the IRS claims that the excise tax not only covers this flight-by-flight usage charge but that it also covers loads of other “reasonably necessary” services covered by the fixed management and overhead fees. These are many and wide ranging: the initial jet share purchase, pilot medical examinations and uniforms, flight and ground crew salaries, maintenance, inspections, service and repair, exterior repainting, interior refurbishing, computerized maintenance programs, plane tie-downs, aircraft storage and hangar space, FAA and FCC paperwork, insurance, claim settlement, human resources, catering, communications between fractional jet owners, and other miscellaneous administrative services. How does Flight Options know which costs and services to treat as taxable and which ones to exclude? Other than the phrase 'reasonably necessary' (which is not in the statute or regulations) and other than the regulations' concession that the excise tax does not apply to parking, deicing, sanitation, dockage, wharfage, “etc.,” the taxpayer (in reality the private tax collector) has nothing to go on.

The oral argument in this case captured the uncertainty. The government’s lawyer, who ably represented his client, was asked whether the excise tax applied to the initial charge for ownership of a fractional share of a jet. That is a one-time charge for each Flight Options fractionalshare owner. It is a 'reasonably necessary' cost for every flight of the jet that subsequently occurs. And yet the IRS has never imposed the tax on this initial transaction. The government’s lawyer acknowledged that the IRS “probably should have” applied the tax to those costs and transactions from the outset. Oral Arg. 17:39 –18:21. The logic of the government’s position indeed would apply this “ticket tax” to every purchase of every vehicle of transportation: all planes, all trains, all automobiles. How strange. Eighteen years after the first year covered by this audit, and fourteen years after Congress eliminated the application of this Page 12excise tax on usage or fixed fees by the fractional-share jet industry, we learn that the largest fee charged by the industry should have been covered. We don’t think the government is right given the operative “for transportation” language that describes this “ticket tax.” But the government’s honest answer covers the predicament these private tax collectors faced in deciding where to draw the line. If “the Government is uncertain about the precise application of the test that it proposes” and if the government has not taken a consistent position in applying it, Burrage v. United States , 571 U.S. 204, 217 – 18 (2014), that suggests the citizen faces a similar predicament.

That’s one conspicuous uncertainty about where the tax starts and stops. Here are a few others. Flight Options employs pilots, ground crews, fleet managers, technicians, painters, HR representatives, and inhouse counsel. Which of these roles are 'reasonably necessary' for a flight and which are optional? Lawyers and office clerks don’t fly planes, but they do handle the mandatory paperwork and regulatory compliance. The same question naturally applies to Flight Options’ ground services: air traffic control, coordination between fractional -jet owners, repairs and refurbishment, and insurance and claim settlement. While some of these services track specific flights and others maintain an ongoing fleet, no responsible jet owner would lightly forgo any of them.

Then there’s the allocation problem when these services support domestic and international flights. Because § 4261 applies only to domestic (and near-domestic) personal travel, a roundtrip ticket to Tokyo and back falls outside the tax. See 26 U.S.C. § 4262(b). Nothing in the statute tells a fractional jet owner who exclusively flies abroad that the tax applies to his monthly fees (or, for that matter, his hourly usage fees) when in reality those fees apply only to untaxed transportation. But what of the fractional jet owner who leaves his plane and its pilots in the hangar (or thereabouts) for a month instead of flying to London? He still pays the monthly fee, and the monthly fee still goes to costs that allow domestic flights. But the monthly fee also purchases the option to fly abroad, and by statute amounts “paid for” foreign flights are untaxed. It would be exceedingly odd if a taxpayer triggered a transportation tax by flying not at all when he avoided the tax altogether by flying some. Surely it makes more sense to say that the tax does not fall on someone who declines to fly at all.

Page 13That leaves the question of whether and how the tax falls on a fractional jet owner who flies abroad some, flies domestic some, and leaves some hours unused. Different taxation regimes, for what it is worth, have answered this sort of question differently. For example: Congress once imposed an excise tax on “amount[s] paid for admissions” to certain events. 26 U.S.C. § 1700 (1940). That approach left open how to tax membership dues for clubs that offered their members, among other things, access to such activities for a per-event fee. See Newland v. United States , 220 F.2d 925, 926 (9th Cir. 1955). The IRS interpreted the statute to apply the tax to membership dues only if their 'chief or sole privilege' was 'a right of admission” to a covered event “on a definite number of occasions.” 26 C.F.R. § 101.2 (1941). Would that be the right approach here if the IRS is right — that is, to look to the chief use of the aircraft that month? That is a lot to ask of someone tasked with responsibility for collecting withholding taxes and burdened with liability for getting it wrong.

The lack of available IRS guidance left Flight Options’ collection obligations “speculative” and far from “precise.” Central Illinois , 435 U.S. at 31. As of 2012, “no court had ever held' that charges for jet memberships and management services incurred a tax under § 4261. Id . at 32. The sole published case noted instead that '[n]o transportation tax was assessed with respect to the monthly fee” collected by a fractional -share jet company. Exec. Jet , 125 F.3d at 1467.

It’s not clear whether the government has authority to remove these uncertainties through regulations as opposed to statutes in this setting. See Carter v. Welles-Bowen Realty, Inc. , 736 F.3d 722, 729 (6th Cir. 2013) (Sutton, J., concurring); see also Esquivel-Quintana v. Lynch , 810 F.3d 1019, 1028 – 30 (6th Cir. 2016) (Sutton, J., concurring in part and dissenting in part), rev’d , 581 U.S. 385 (2017). What matters for today’s purposes is that the IRS never clarified how this tax applies to the fractional-share jet industry. Central Illinois , 435 U.S. at 32. No regulation ever specified which fixed charges were “for transportation,” leaving taxpayers to figure which payments were 'similar' to non -transportation charges such as 'parking,' '[de]icing,' and 'sanitation,' 26 C.F.R. § 49.4261-8(f)(5) (2012), and which were not. Nor did the IRS ever regulate how to treat membership fees or otherwise allocate fixed expenses, a failure all the more jarring because Congress directed it to provide an allocation scheme. See 26 U.S.C. § 4263(c);Page 14 see also 26 C.F.R. §§ 101.2 (1941) (distinguishing admissions and membership fees), 1.512(a)1(c) (prescribing the allocation of fixed expenses).

The IRS’s informal rulings did no better. The IRS wrote two conflicting technical advice memoranda on the subject. One, issued in 1992 to NetJets’ predecessor company, applied the § 4261 tax only to the hourly fee. The other, issued in 2004 to Bombardier, applied the tax to monthly fees but without a prescribed allocation. See NetJets Large Aircraft, Inc. v. United States , 80 F. Supp. 3d 743, 756 (S.D. Ohio 2015); IRS Tech. Adv. Mem. 2004 – 42 – 5048. The IRS knew about the confusion, as one of its 'analyst[s] suggested that the law was unclear because the IRS Office of Appeals had' (as the government puts it) 'incorrectly ruled in the favor of taxpayers.' Appellee's Br. 51 (quoting R.74 -2). Although Flight Options and other industry participants requested guidance from the IRS, no revenue ruling ever addressed fractional jet programs or membership and management fees. NetJets , 2015 WL 7784925, at *8 – 10 (noting lack of relevant revenue rulings); id. at *11 (concluding that IRS failed to provide 'precise and not speculative' guidance regarding monthly management fees). Instead of alleviating confusion, the IRS’s enforcement efforts added to it. Although the government read § 4261 to apply to charges “reasonably necessary” for air transportation, it never practiced what it preached. For the first twenty years of the fractional jet industry, the IRS did not assess taxes on management fees. See Exec. Jet , 125 F.3d at 1467 –69. When the IRS’s position changed in 2004, it repeatedly declined to bring enforcement actions against fractional-share jet managers for uncollected taxes on management fees and conceded any amounts assessed when brought to court. The IRS nonetheless purported to rely on these company-by-company audits to provide industry guidance despite recognizing that “specific advice [was] needed to clarify the law” on management fees. R.74-2.

Making matters more burdensome and imprecise, the IRS adopted a regulation requiring thirdparty tax collectors to levy 'separable' 'charges for nontransportation services' and “charge[s] for transportation.” 26 C.F.R. § 49.4261-2(c). That approach is fine in the abstract. But it is exceedingly difficult to justify when the Service fails to provide guidance about the charges that fall on the taxable or non-taxable side of the line and for that matter the collectible or non-collectible side of the line — and when the Service raises the stakes by imposing a Page 15retroactive excise tax on the “full amount” collected for all nontransportation services if charged alongside any service incorrectly identified as not “for transportation.” Id. In the absence of a workable or understandable theory of which fixed costs are “for transportation,” the IRS had no right to thrust the burden of definition and collection on Flight Options. The point of the clarity canons for taxpayers and tax withholders is to avoid this problem. It’s not the taxpayer’s duty to devise a workable interpretation of § 4261, to apply it to the gamut of services it provides, hope that it got everything right, and pray it does not face serious retroactive liability from one direction or the other. Such 'retroactive' liability raises the exact concerns that motivate the third-party collector canon. Central Illinois , 435 U.S. at 32 n.12. The government, not Flight Options, carries the burden of precisely identifying which payments incur a tax.

By adopting a theory lacking “explicit standards,” the IRS guaranteed that § 4261 would be enforced 'on an ad hoc and subjective basis, with the attendant dangers of arbitrary and discriminatory application.” Grayned v. City of Rockford , 408 U.S. 104, 108 – 09 (1972). Recall that a third-party tax collector risks personal liability for undercollection, a risk paired with the general threat of criminal liability for tax underpayment. See 26 U.S.C. §§ 6672, 7203. It should come as no surprise that Flight Options and the rest of the fractional-share jet industry repeatedly complained of discriminatory enforcement practices and asked for clearer guidance on which fixed fees count as “for transportation.” See R.156 at 22 – 23; see also Bombardier , 831 F.3d at 279 – 80.

We do not permit the government to cast a large and indeterminate net, then let prosecutors and accountants decide who faces liability after the fact. Because Flight Options lacked precise and not speculative guidance to collect taxes primarily owed by its clients, the government may not now hold it secondarily liable. Central Illinois , 435 U.S. at 33.

The government resists this conclusion on several grounds, each unavailing. It claims that the statute clearly imposes a collection obligation on Flight Options for fixed overhead and management fees. The phrase 'for transportation,' says the government, permits a broad meaning if taken in its “most general sense,” encompassing everything up to and including “the Page 16entire body of services rendered by a carrier.” Appellee’s Br. 24– 25 (quotations omitted). But no canon of interpretation dictates that we look for the “most general” meaning of a tax statute and apply it in favor of the government. If anything, the pro-taxpayer canon, when invoked, applies the opposite rule to tax laws. See OfficeMax, Inc. , 428 F.3d at 594 (collecting cases). Because the statute does not define transportation, it takes its 'common meaning,' United States v. Moses , 137 F.3d 894, 899 (6th Cir. 1998), a meaning much narrower than an 'entire body of services' or for that matter any 'reasonably necessary' service connected to air transportation. If the government were right, some of its own acknowledged exclusions — parking, de-icing, sanitation, the purchase of the plane in full — would fall within the tax.

The government claims that Flight Options' management and other services fit under “waiting time,” which § 4262 lists as part of “transportation.” It points to Shell Oil v. United States , a 1979 Court of Claims case deciding that payments for “on -demand” helicopter service paid for “waiting time” and thus fit within § 4261. 221 Ct. Cl. 296, 300 (1979). But the sense of “waiting time” in § 4262(d) — that “‘transportation’ includes layover or waiting time”— speaks to waiting times for particular flights, not to the general management services provided by Flight Options for all flights and for all customers.

The government turns to § 4263(d), which says that the ticket tax 'appl[ies] to any amount paid within the United States for transportation of any person by air unless the taxpayer establishes . . . that the transportation is not transportation in respect of which tax is imposed by section 4261.” 26 U.S.C. § 4263(d); see also 26 C.F.R. § 49.4261-4(a) (similar). This section creates only a record-keeping requirement, and no one claims that Flight Options failed to satisfy it, whether with respect to some amounts or “any amount.” As for its potential bearing on Flight Options’ withholding obligation, the provision takes a circular route back to the question at hand: Are fixed overhead fees, as opposed to usage fees, charged “for transportation” and thus eligible for this “ticket tax”? Section 4263 simply re -asks the question; it does not answer it.

The government also turns to an adjacent regulation. It claims that Flight Options’ fixed management and membership fees were analogous to “prepaid orders” for transportation under the regulations, with the hourly fees serving in essence as 'additional charges' required to redeem prepaid transportation. 26 C.F.R. § 49.4261-7(c), (f). The analogy does not work. APage 17 prepayment buys a specific service in advance. See Webster's Third New International Dictionary, supra , 1790 –91 (defining “prepaid expense” as “a deferred charge” and “prepayment” as “payment in advance”); see also id. at 591 (defining 'deferred charge' as an 'expense . . . incurred prior to the fiscal period to which it applies'). The management and membership fees gave clients the right to purchase flight hours, but those clients still needed to pay a usage charge for each specific flight. When clients prepaid their usage fees, Flight Options collected the tax. When they did not, Flight Options did not collect the tax.

The government points to IRS revenue rulings saying that any payment would incur the tax if “reasonably necessary” for air transportation. Rev. Rul. 80 -31 (Feb. 4, 1980); Rev. Rul. 2006-52 (Oct. 23, 2006). We see no coherent way to apply this guidance, given the many costs the government has never taxed, to determine whether a fixed cost constituted a payment “for transportation' or how to allocate such costs between taxable and nontaxable transportation. Neither, apparently, did the IRS, as it admits that it applied the “reasonably necessary” test only to variable charges purchased in connection with a ticket on a flight-by-flight basis. See Appellee’s Br. 30– 31.

We also doubt whether the “reasonably necessary” standard correctly interprets § 4261 in the first instance. It equates “amounts paid for transportation” with “amounts paid reasonably necessary for transportation,” using two new words that Congress knew how to use on its own but never did. See, e.g. , 26 U.S.C. §§ 993(b)(4); 5301(a)(2); 7610(a)(2). Plus, a 'reasonably necessary” test has few stopping points, sweeping in all sorts of costs— say the cab ride to the airport or the cost of the building where people wait for a flight —that fall outside “transportation by air.” The standard also conflicts with the IRS’s own regulations, which treat “[de]icing” and “parking” services— other essential preconditions to flight — as non-transportation charges.

For like reasons, we see no role for Skidmore respect. It applies “only to the extent that” the revenue rulings have “the ‘power to persuade.’” Christensen v. Harris County , 529 U.S. 576, 587 (2000) (quoting Skidmore v. Swift & Co. , 323 U.S. 134, 140 (1944)). The 'one -page analysis in [these] revenue ruling[s]” and the absence of a limiting principle do “not contain the traditional hallmarks for receiving deference.” OfficeMax , 428 F.3d at 594 – 95.

Page 18The government worries that we overread Central Illinois . It argues that Central Illinois ’ requirement that withholding obligations be “precise and not speculative” prohibits nothing more than 'retroactive application of withholding obligations.' Appellee's Br. 42. But lack of statutory notice of a withholding obligation before the relevant taxable event is precisely what made the government's position 'somewhat retroactive in character' in Central Illinois . 435 U.S. at 32 n.12. And it’s precisely what would do the same here. Every court to consider the question has read Central Illinois to require the IRS to provide 'precise and not speculative' guidance before imposing a withholding tax. See, e.g. , W. Rsrv. Acad. , 801 F.2d at 251, adopting 619 F. Supp. 400 – 01; Bombardier , 831 F.3d at 278 – 89; Univ. of Chi. v. United States , 547 F.3d 773, 784 (7th Cir. 2008); HB & R , 229 F.3d at 691 – 92; Am. Airlines , 204 F.3d at 1111 – 12; Gen. Elevator Corp. v. United States , 20 Cl. Ct. 345, 353 (1990).

Falling back, the government claims that the 2004 technical advice memorandum applying the tax to management fees left “no doubt about the IRS’s position.” Appellee’s Br. 46 – 47. But the question is not what an IRS agent or an IRS technical advice memorandum thinks the law is. It is whether a binding statute or regulation applies to the transaction. A memorandum created by a party to a case no more establishes the law than the absence of a memorandum undercuts the clear meaning of the law. That’s why Central Illinois looked to the “statutory requirement,” binding agency “regulations” and “ruling[s],” and “judicial decisions” to establish the requisite “precis[ion].” Central Illinois , 435 U.S. at 31 – 32. The government also ignores the actual law, which provides that neither the IRS nor taxpayers may rely on tax advice memoranda issued to other companies. See 26 U.S.C. § 6110(k)(3); 26 C.F.R. § 601.601(d)(1) (“No unpublished ruling or decision will be relied on, used, or cited by any officer or employee of the Internal Revenue Service as a precedent in the disposition of other cases.'); cf. Bombardier , 831 F.3d at 279 – 83 (applying technical advice memorandum ruling on management fees applicable only to Bombardier).

The “Audit Technique Guide” sent by an IRS field agent to Flight Options in 2008 has similar flaws and limitation. R.1611 at 2. As the guide warned, '[t]his document is not an official pronouncement of the law or the position of the Service and can not be used, cited, or relied upon as such.” R.161 -1 at 2. Although the guide concluded that management fees include Page 19payments for “taxable transportation,” it did not identify which specific fixed costs counted as “for transportation” or how to allocate those costs. The guide at most notified Flight Options of a need to withhold taxes for an indefinite and undefined part of its business. That is no more notice of an obligation than a speeding ticket issued for violating some speed limit some place some time ago.

The government points to a 2006 internal memo authored by tax counsel working for Raytheon, Flight Options' then -parent company. Because the memo concluded that Flight Options had a “potential liability [of] zero to $69.2” million on management fees, R.73-4 at 9, the government claims that Flight Options had notice of some potential liability. This argument misstates the law and the facts. The law: The adequacy of taxpayer notice does not depend on Flight Options’ subjective and speculative thinking about what the law might be, any more than it depends on what the IRS thinks the law is. Notice, to repeat, turns on objective 'statutory requirement[s],” binding agency “regulations” and “ruling[s],” or “judicial decisions.” Central Illinois , 435 U.S. at 31 -32. The facts: Flight Options' awareness that it had 'potential' tax exposure ranging from zero to $70 million shows that its obligations were anything but “precise” and 'non speculative.'

For these reasons, we reverse.

Page 20MURPHY, Circuit Judge, concurring. I agree with Chief Judge Sutton's excellent majority opinion that the best reading of the Tax Code supports Flight Options’ position in this case. And if there were any doubt about the proper interpretation, I also agree that the relevant canons of construction would confirm the company's reading. I write further to flag the confusing conflict in the Supreme Court’s caselaw on the governing canons for interpreting the federal tax laws.

The Court has followed a winding path when it comes to these canons of interpretation. It started with a protaxpayer canon. In 1873, the Court opined that it must resolve any “doubt as to the liability of an instrument to taxation” “in favor of [an] exemption” because “a tax cannot be imposed without clear and express words for that purpose.” United States v. Isham , 84 U.S. 496, 504 (1873) (quoting Gurr v. Scudds , 11 Exch. 190, 192, 156 Eng. Rep. 798, 799 (1855) (Pollock, C.B.)). Over the next six decades, the Court issued many cases reciting this protaxpayer canon. See, e.g. , Old Colony R.R. Co. v. Comm’r , 284 U.S. 552, 561 – 62 (1932); United States v. Merriam , 263 U.S. 179, 188 (1923); Shwab v. Doyle , 258 U.S. 529, 536 – 37 (1922); Gould v. Gould , 245 U.S. 151, 153 (1917); Benziger v. United States , 192 U.S. 38, 55 (1904); Eidman v. Martinez , 184 U.S. 578, 583 (1902); Am. Net & Twine Co. v. Worthington , 141 U.S. 468, 474 (1891); Hartranft v. Wiegmann , 121 U.S. 609, 616 (1887). By 1932, the Court called it “elementary that tax laws are to be interpreted liberally in favor of taxpayers” and that “[d]oubts must be resolved against the government[.]” Miller v. Standard Nut Margarine Co. of Fla. , 284 U.S. 498, 508 (1932).

Yet something changed around the New Deal. Within a decade, the Court switched from this longstanding taxpayer-favoring canon to a novel taxpayer-disfavoring one. See Erwin N. Griswold, An Argument Against the Doctrine that Deductions Should Be Narrowly Construed as a Matter of Legislative Grace , 56 Harv. L. Rev. 1142, 1142 – 45 (1943). In 1934, the Court opined that tax “deductions” were a matter of “legislative grace” and that it would interpret a tax Page 21provision to establish such a deduction only if the provision included “clear” language to that effect. New Colonial Ice Co. v. Helvering , 292 U.S. 435, 440 (1934). In the ensuing decades, the Court came to call it a “familiar rule” that taxpayers bear the “burden” to “clearly” prove that they fall within a tax exemption. Interstate Transit Lines v. Comm’r , 319 U.S. 590, 593 (1943); see Helvering v. Nw. Steel Rolling Mills , 311 U.S. 46, 49 (1940); Deputy v. du Pont , 308 U.S. 488, 493 (1940); White v. United States , 305 U.S. 281, 292 (1938). Even in more recent times, the Court has sometimes said things like 'exemptions from taxation are to be construed narrowly.” Mayo Found. for Med. Educ. & Rsch. v. United States , 562 U.S. 44, 59 – 60 (2011) (quoting Bingler v. Johnson , 394 U.S. 741, 752 (1969)); see INDOPCO, Inc. v. Comm’r , 503 U.S. 79, 84 (1992); United States v. Centennial Sav. Bank FSB , 499 U.S. 573, 583 (1991).

So which canon governs here? For the reasons explained by the majority opinion, the pro-taxpayer canon should at least apply where, as here, the government seeks to hold a third party liable for failing to collect taxes owed by someone else. See Cent. Ill. Pub. Serv. Co. v. United States , 435 U.S. 21, 31 – 32 (1978). Yet, apart from this context of third-party tax collectors, the tension between these competing canons will remain. The Supreme Court has no authoritative opinion jettisoning the older pro-taxpayer canon or justifying the newer progovernment canon.

My review of this caselaw leads me to reach two tentative conclusions about these canons. To start, when uncertainty exists over which canon to follow, courts should adhere to the pro-taxpayer canon because it has a stronger foundation. Next, and nevertheless, this protaxpayer canon does not allow courts to depart from the best reading of a tax law. Like the rule of lenity, it instead allows us only to choose between two equally plausible views of the law.

Pro-Taxpayer Canon v. Pro-Government Canon . For several reasons, I would generally follow the canon interpreting tax obligations narrowly over the competing canon interpreting them broadly whenever the law is unclear over which applies. First , the pro-taxpayer canon has a much firmer 'historical pedigree' than does the canon favoring the government. Biden v. Nebraska , 600 U.S. 477, 509 (2023) (Barrett, J., concurring). Some 'two hundred reported decisions' and 'thousands of taxpayers' briefs' had invoked this pro -taxpayer canon by the 1930s. See Griswold, supra , at 1142. True, my research tracing the Supreme Court’s use of the Page 22canon dates only to 1873 —decades after our founders established the “judicial Power” in Article III. U.S. Const. art. III, § 1; see Isham , 84 U.S. at 504. But other sources show that it has deeper roots.

To start, authorities from the 1870s recognized that the “state[]” and “federal courts” had long followed the pro-taxpayer canon. Thomas M. Cooley, A Treatise on the Law of Taxation 200, 202 (1876); see, e.g. , Powers v. Barney , 19 F. Cas. 1234, 1234 (C.C.S.D.N.Y. 1863); Moseley v. Tift , 4 Fla. 402, 403 – 04 (1852); Sewall v. Jones , 26 Mass. 412, 421 (1830). Justice Story, for example, invoked it several times. In one case involving a tariff, he explained that Congress must enact laws imposing taxes “in a clear and determinate manner” and that courts should never rely on 'doubtful interpretations' to impose 'duties' on citizens. Adams v. Bancroft , 1 F. Cas. 84, 85 (C.C.D. Mass. 1838). In a similar decision, he said that courts should “[i]n every case . . . of doubt” construe a tax law “strongly against the government, and in favor of the subjects[.]' United States v. Wigglesworth , 28 F. Cas. 595, 597 (C.C.D. Mass. 1842). Likewise, thenJustice Kent interpreted an even earlier federal stamp act 'strictly' to exclude items that did not clearly fall within its terms. Conroy v. Warren , 3 Johns. Cas. 259, 264 – 65 (N.Y. 1802).

According to Justice Cooley, a similar rule of construction had 'always prevailed in England.” Cooley, supra , at 201 & n.2. By 1831, a treatise described it as “well settled” “that every charge upon the subject must be imposed by clear and unambiguous language.' 2 Fortunatus Dwarris, A General Treatise on Statutes 479 (1831). Six years earlier, a judge reiterated this “well settled rule” when refusing to treat a real -estate transfer as taxable. Denn v. Diamond , 4 B. & C. 243, 245, 107 Eng. Rep. 1049, 1050 (K.B. 1825) (Bayley, J.). Another judge explained even earlier that courts 'should give a liberal construction to words of exception' in a tax 'charged' on a 'subject' because the legislature must 'fairly mark[] out' these taxes. Warrington v. Furbor , 8 East 242, 245, 103 Eng. Rep. 334, 335 – 36 (K.B. 1807) (Ellenborough, C.J.).

Second , the pro-taxpayer canon resembles an even more established rule: the rule of lenity. See Antonin Scalia & Bryan A. Garner, Reading Law 299 – 300, 359 – 60 (2012). As Blackstone explained before the founding, '[p]enal statutes must be construed strictly.' 1 St.

Page 23George Tucker, Blackstone's Commentaries 87 (1803). And, as Chief Justice Marshall explained after it, this rule “is perhaps not much less old than construction itself.” United States v. Wiltberger , 18 U.S. 76, 95 (1820). To be sure, the Court has rejected the notion that it should treat all revenue provisions as “penal laws” subject to a strict construction. Rankin v. Hoyt , 45 U.S. 327, 332 (1846); see United States v. Hodson , 77 U.S. 395, 406 (1870); Cooley, supra , at 203 – 05.

Still, tax laws share some attributes with criminal laws. Cf. Pasquantino v. United States , 544 U.S. 349, 360 –61 (2005). The ability to tax represents one of “the government’s greatest powers.' FCC v. Consumers’ Rsch. , 606 U.S. 656, 723 (2025) (Gorsuch, J., dissenting); see Learning Res., Inc. v. Trump , 146 S. Ct. 628, 637 – 38 (2026). Tax laws coerce parties and “affect[] private rights” in ways similar to criminal laws. City of Washington v. Pratt , 21 U.S. 681, 683 (1823). So those who objected to the federal taxing power noted that, “[b]y means of taxes, the government may command the whole or any part of the subject’s property.” Federal Farmer, Letter III (Oct. 10, 1787), in 2 The Founders’ Constitution 412– 13 (Philip B. Kurlan & Ralph Lerner eds., 2000). And it is unsurprising that the “power to tax” has long been said to include “the power to destroy[.]” McCulloch v. Maryland , 17 U.S. 316, 431 (1819).

The pro-taxpayer canon thus follows from the same separation-of-powers concerns that underlie the rule of lenity. Just as Congress must clearly delegate its power to allow the Executive Branch to create a crime, see United States v. Eaton , 144 U.S. 677, 687 – 88 (1892), so too it must clearly delegate its power to allow the Executive Branch to impose (at the least) an internal tax, see Learning Res. , 146 S. Ct. at 637 – 39; cf. id. at 687 – 88 (Thomas, J., dissenting). And just as judges may not create common-law crimes, see United States v. Hudson , 11 U.S. 32, 34 (1812), so too they may not impose common-law taxes, see Missouri v. Jenkins , 495 U.S. 33, 65 – 67 (1990) (Kennedy, J., concurring in part and concurring in the judgment). The rule of lenity and the pro-taxpayer canon ensure that Congress (not the courts) retain these constitutional powers. See Amy Coney Barrett, Substantive Canons and Faithful Agency , 90 B.U. L. Rev. 109, 133 – 34 (2010); Wooden v. United States , 595 U.S. 360, 391 (2022) (Gorsuch, J., concurring in the judgment).

Page 24Besides, many (perhaps most) tax obligations trigger civil or even criminal penalties. One need look no further than this case. And many other examples exist. See, e.g. , United States v. Thompson/Ctr. Arms Co. , 504 U.S. 505, 517 – 18 (1992) (plurality opinion); Comm’r v. Acker , 361 U.S. 87, 91 (1959). The pro-taxpayer rule of lenity would apply in all these tax contexts. See Bittner v. United States , 598 U.S. 85, 102 – 03 (2023) (Gorsuch, J., opinion). Indeed, the Supreme Court has many cases applying this canon to tax laws that imposed the forfeiture of property as a penalty for a taxpayer’s failure to pay taxes owed on the property. See Bennett v. Hunter , 76 U.S. 326, 335 – 36 (1869); Dubois v. Hepburn , 35 U.S. 1, 22 – 23 (1836); United States v. Eighty-Four Boxes of Sugar , 32 U.S. 453, 462 – 63 (1833). These decisions found these forfeiture laws “highly penal” and so interpreted them to favor the taxpayer’s right to redeem the property by paying the past-due taxes. Bennett , 76 U.S. at 336.

Granted, the Court has another line of older cases (consistent with the utter confusion in this area) suggesting that even tax penalties 'are construed less narrowly' than criminal penalties. United States v. Ryan , 284 U.S. 167, 172 (1931); United States v. Stowell , 133 U.S. 1, 12 (1890); Hodson , 77 U.S. at 406. This precedent makes little sense to me because it existed at the same time the Court applied the pro-taxpayer canon to all tax laws (not just tax penalties). See Gould , 245 U.S. at 153; Isham , 84 U.S. at 504. The precedent also conflicts with the line of cases cited above that gave a “milder construction” to forfeiture laws in the tax context. Bennett , 76 U.S. at 336. And it conflicts with the Court’s modern decisions favoring the taxpayer when it comes to the interpretation of tax penalties. See Acker , 361 U.S. at 91; see also Bittner , 598 U.S. at 102 – 03 (Gorsuch, J., opinion). So I would view this line of cases narrowly as rejecting only a strong-form version of the pro-taxpayer canon (of which I will speak more about below).

Third , the Supreme Court's cases ignoring or rejecting the pro -taxpayer canon in the 1930s did not claim that the canon lacked a historical foundation. Rather, these cases departed from the canon on grounds that the Court would view with skepticism today. See Assaf Likhovski, The Duke and the Lady: Helvering v. Gregory and the History of Tax Avoidance Adjudication , 25 Cardozo L. Rev. 953, 977 – 83 (2004). I will address the three main criticisms in turn.

Page 25Criticism One: Justice Cardozo began the dismantling of the pro-taxpayer canon in a 1932 decision seeking to prevent tax avoidance. See Woolford Realty Co. v. Rose , 286 U.S. 319, 329 – 30 (1932). In Woolford Realty , the Court refused to adopt the taxpayer's reading of a tax law because of the “mischiefs” the reading would engender. Id. at 329. According to the Court, the 'mind rebel[led]' against the taxpayer's reading because it would allow a 'prosperous corporation' to 'wipe out' its tax obligations. Id. at 329 – 30. So the Court ignored the protaxpayer canon and adopted a canon against tax avoidance instead. It reasoned that “[e]xpediency” (that is, policy) “may tip the scales when arguments are nicely balanced .” Id. at 330 (emphasis added). This logic adhered to Justice Cardozo’s general views of interpretation, which called for judges to rely on wise policy to fill gaps in ambiguous laws. See Benjamin N. Cardozo, The Nature of the Judicial Process 66 –67 (1921). And the case started the Court’s switch from a “formalist, strict approach” to a purpose -laden “‘equitable’ approach.” Likhovski, supra , at 980 (citation omitted).

The problem? Over the next few decades, the Court continued down this purpose-overtext path in all contexts (not just the tax context). See Gun Owners of Am., Inc. v. Bondi , 168 F.4th 813, 823 (6th Cir. 2026). But the momentum has since swung back sharply the other way. See id. ; Scalia & Garner, supra , at 13. Nowadays, the Court has returned to the more formalist approach to interpretation. It recognizes that “no statute yet known pursues its stated purpose at all costs” and that courts cannot rely on abstract notions of purpose to clarify ambiguities in the law. Stanley v. City of Sanford , 606 U.S. 46, 58 (2025) (quoting Henson v. Santander Consumer USA Inc. , 582 U.S. 79, 89 (2017)). As a result, Woolford Realty 's consequentialist call to interpret ambiguities to prohibit tax avoidance sits uncomfortably next to the Court's modern precedents.

Criticism Two: In a decision the next year, the Court criticized the pro-taxpayer canon because an interpretation of a tax provision often will help some taxpayers but harm others. See Burnet v. Guggenheim , 288 U.S. 280, 286 (1933). True, this fact would provide a basis to set aside the canon in a specific case where no single reading would benefit taxpayers. But the fact did not provide a basis to categorically reject the canon. An analogy shows why. The Court has made the same point for the rule of lenity: sometimes an interpretation will help some criminal Page 26defendants but harm others. See Brown v. United States , 602 U.S. 101, 122 – 23 (2024); United States v. Delaine , 171 F.4th 880, 888 (6th Cir. 2026). Under the Court’s logic in Burnet , this fact should have led it to altogether ditch the rule of lenity. But the Court did no such thing. Rather, it held more narrowly that the rule of lenity does not apply in these specific situations. See Brown , 602 U.S. at 122 – 23. The same logic could have extended to the pro-taxpayer canon.

Criticism Three: The Court in a 1938 decision criticized the pro-taxpayer canon as an abdication of the judicial role. See White , 305 U.S. at 292. When refusing to apply the canon, the Court reasoned that it was “the function and duty of courts to resolve doubts” in statutes. Id. It added that 'doubts which may arise upon a cursory examination' of the relevant provisions might 'disappear' when read against the entire statutory scheme and 'in the light of their legislative history.” Id. This reasoning in part fights a strawman and in part rests on a mistake. Yes, courts should not simply read a tax provision in isolation, find it ambiguous, and call it a day. Rather, courts should determine whether the provision has a clear meaning by applying “all of the traditional tools of statutory interpretation[.]” Shular v. United States , 589 U.S. 154, 166 – 67 (2020) (Kavanaugh, J., concurring). Only if doubt remains at this point may they invoke the canon. See id. So the canon no more requires courts to “abdicate[]” their traditional role than does the rule of lenity. White , 305 U.S. at 292. But no, the courts should not resort to “legislative history” to clarify the meaning of a text that it finds ambiguous after exhausting all the traditional tools of interpretation. Id. Instead, 'the next step' should be to invoke the canon — just as it should be in the rule-of-lenity context. Wooden , 595 U.S. at 395 (Gorsuch, J., concurring in the judgment); see United States v. R.L.C. , 503 U.S. 291, 308 – 10 (1992) (Scalia, J., concurring in the judgment).

In sum, the Court refused to follow the pro-taxpayer canon in the 1930s either for outdated reasons or for legitimate reasons that do not justify a categorical disavowal. Yet these cases did not expressly overrule the canon. So the Justices are free to revive it in a proper case. Cf. United Dominion Indus., Inc. v. United States , 532 U.S. 822, 838 – 39 (2001) (Thomas, J., concurring).

Fourth , and finally, the competing pro-government canon has a dubious pedigree. It arose from a single (citationless) sentence in a 1934 tax case: 'Whether and to what extent Page 27deductions shall be allowed depends upon legislative grace; and only as there is clear provision therefor can any particular deduction be allowed.' New Colonial Ice , 292 U.S. at 440; see Griswold, supra , at 1143 – 44. The Court then repeatedly cited New Colonial Ice to claim that it should narrowly construe tax exemptions. See INDOPCO , 503 U.S. at 84; Interstate Transit Lines , 319 U.S. at 593; Helvering , 311 U.S. at 49 & n.7. Yet courts had previously invoked the competing protaxpayer canon even when interpreting an 'exemption' or 'exception' to taxation. Isham , 84 U.S. at 504 (citation omitted); Warrington , 103 Eng. Rep. at 335 – 36. And the Court did not explain the switch in New Colonial Ice or any later decision. As a result, I fear that this 'unexamined assumption[]' in New Colonial Ice has 'becom[e], by force of usage, unsound law.” McCormick v. United States , 500 U.S. 257, 280 (1991) (Scalia, J., concurring).

Consider an analogy to the Fair Labor Standards Act. There, too, the Supreme Court once said that courts should 'narrowly construe[]' the exemptions to its minimum -wage and maximum-hour requirements. Arnold v. Ben Kanowsky, Inc. , 361 U.S. 388, 392 (1960); see A.H. Phillips, Inc. v. Walling , 324 U.S. 490, 493 (1945). Yet because the law did not include any “textual indication” of this narrow -construction canon, the Court has since rejected the canon as an unreliable “guidepost” to interpreting that law. Encino Motorcars, LLC v. Navarro , 584 U.S. 79, 88 – 89 (2018) (quoting Scalia & Garner, supra , at 363).

It is likely only a matter of time before the Court recognizes the same problem in this tax context. Indeed, the treatise on which the Court relied in Encino Motorcars specifically mentioned the 'false notion' that ' tax exemptions' 'should be strictly construed.' Scalia & Garner, supra , at 359 (emphasis added). As this treatise explained, one must be careful to differentiate this dubious canon from other longstanding ones. See id. at 359 – 62. For example, taxpayers sometimes seek an exemption from state taxation on the ground that a federal law preempts the tax. See Fla. Dep’t of Revenue v. Piccadilly Cafeterias, Inc. , 554 U.S. 33, 48 – 50 (2008). And they sometimes claim that a State has entered a contract not to tax them and attempt to enforce this purported contract term under provisions of the U.S. Constitution (such as the Contracts Clause). See Providence Bank v. Billings , 29 U.S. 514, 561 (1830). It is established in these other contexts that courts should interpret the federal law (or purported contract term) narrowly in favor of the States’ power to tax. See Fla. Dep’t of Revenue , 554 U.S. at 48 – 50;Page 28 Providence Bank , 29 U.S. at 561; see also Scalia & Garner, supra , at 360 – 61. But these rules say nothing about how to interpret a federal exemption in a federal tax. Instead, the pro-taxpayer canon that the Court had long applied before the 1930s likely extended to this context.

Strong v. Modest Canon . Yet one should not overread what I have said. In all likelihood, the protaxpayer canon will 'rarely' matter to the outcome. Wooden , 595 U.S. at 377 (Kavanaugh, J., concurring). As Justice Barrett has explained, substantive canons come in “strong” and “modest” forms. Biden , 600 U.S. at 508 (Barrett, J., concurring). A “strong -form canon” tells a court to accept a plausible reading of a text even if the court believes the chosen reading is not the best one. Id. Under the canon of constitutional avoidance, for example, a court will adopt a 'possible' or 'plausible' reading so that it may sidestep the constitutional question that the “best” reading would require it to resolve. Arizona v. Inter Tribal Council of Ariz., Inc. , 570 U.S. 1, 18 (2013); Clark v. Martinez , 543 U.S. 371, 381 (2005). A 'modest' canon, by contrast, does less work. Biden , 600 U.S. at 508 (Barrett, J., concurring). It acts only as a tiebreaker when choosing 'between equally plausible interpretations of a statute.' Id. Under the rule of lenity, for example, courts should accept the interpretation that favors the criminal defendant only if they still have reasonable doubts about the best reading after exhausting all the traditional tools of interpretation. See Ocasio v. United States , 578 U.S. 282, 295 n.8 (2016). Because, however, most statutes will have a “single, best meaning,” courts will have no need to rely on these modest canons in most cases. Loper Bright Enters. v. Raimondo , 603 U.S. 369, 400 (2024).

The pro-taxpayer canon falls within the modest camp and performs a limited tie-breaking role. The historical authorities prove this point. Take Justice Cooley. When canvassing the caselaw's somewhat conflicting language on this canon, he advocated for a 'just and safe medium' between an overly strict reading that would exclude taxpayers who fell within the statute’s plain meaning and an overly liberal one that would ensnare taxpayers who fell outside that meaning. Cooley, supra , at 205 – 06. In other words, he favored a narrow presumption that courts treat a tax law as covering 'everything' that fell within its text 'without ambiguity or doubt[.]' Id. at 208.

Page 29Or take Justice Story. He equivocated over whether tax laws should be 'construed strictly” like penal laws. Compare Adams , 1 F. Cas. at 85 (yes), with United States v. Breed , 24 F. Cas. 1222, 1222 (C.C.D. Mass. 1832) (Story, J.) (no). But he remained steadfast that these laws should “be construed according to the true import and meaning of their terms,” Breed , 24 F. Cas. at 1222, to cover only what they “expressly and clearly” encompass, Wigglesworth , 28 F. Cas. at 597. Notably, apart from his equivocation over whether to strictly construe the tax laws, Justice Story used similar language to describe this canon as he did to describe the rule of lenity. Cf. United States v. Winn , 28 F. Cas. 733, 734 (C.C.D. Mass. 1838) (Story, J.). His logic would not authorize courts to treat the pro-taxpayer canon as a strong-form canon that could displace the best reading of a tax law. But it would allow courts to treat the pro-taxpayer canon as a modest one that requires them to resolve any “doubt” that remains after examining a law’s text “in favor of the subjects” rather than the government. Wigglesworth , 28 F. Cas. at 597.

Lastly, the Supreme Court cases applying this pro-taxpayer canon from 1873 to the 1930s followed the same approach. The Court reasoned that the canon applied only when “the words” that Congress used had a 'doubtful' meaning. Merriam , 263 U.S. at 188. The canon thus provided no basis to depart from “the clear import of the language used” even if the Court could describe another meaning as plausible. Id. So “when the language [was] clear,” the Court set this canon of construction aside. Treat v. White , 181 U.S. 264, 267 (1901). If courts ruled for the taxpayer even when the plain language triggered the tax, they would not be “resolv[ing] a doubt in [the taxpayer's] favor' but 'say[ing] that the statute does not mean what it means.' Helvering v. Stockholms Enskilda Bank , 293 U.S. 84, 94 (1934).

Just because the canon might not resolve many cases today, though, should not lead courts to underestimate its value. We must place the canon in the context of the interpretive approach that courts followed when they adopted it — not the approach that the Supreme Court follows today. Consider the English experience. English courts long practiced the “equity of the statute” by which they would extend statutes to cover conduct that fell outside their plain terms if the extension served the statute’s purpose. John F. Manning, Textualism and the Equity of the Statute , 101 Colum. L. Rev. 1, 8 (2001) (citation omitted). According to these courts, the protaxpayer canon barred this practice in the tax context. As Lord Cairns explained, “if the Crown, Page 30seeking to recover the tax, cannot bring the subject within the letter of the law, the subject is free, however apparently within the spirit of the law the case might otherwise appear to be.' Partington v. Att’y Gen. , L. R. 4 H. L. 100, 122 (1869) (Cairns, L.). That is, courts could not follow an “equitable construction” to cover taxpayers who fell outside “the words of the statute” even if they fell within its purpose. Id. This mode of interpretation has existed in this country too. See, e.g. , Holy Trinity Church v. United States , 143 U.S. 457, 459 (1892); see also Gun Owners , 168 F.4th at 823. So the canon can serve a similar purpose for federal courts: to prevent those courts from extending a tax law’s reach “in furtherance of the evident purposes and policy of the statute[.]” United States v. Kimball , 26 F. Cas. 782, 785 (C.C.D. Mass. 1844).

Indeed, this limitation might have more importance than one would think even in this textualist age. To this day, the taxing authorities sometimes try to prohibit a tax-avoidance practice that the Tax Code’s text permits on the ground that the practice falls within the general spirit of the relevant tax. See Summa Holdings, Inc. v. Comm’r , 848 F.3d 779, 781 – 82 (6th Cir. 2017). But the protaxpayer canon makes clear that those authorities may not seek to penalize “a Codecompliant transaction in the service of general concerns about tax avoidance.” Id. at 787. Over a hundred years ago, we said that the taxing authorities may not rely on “mere implication from analogies” to tax conduct that does not “plainly” fall within the relevant tax law. United States v. Mullins , 119 F. 334, 337 (6th Cir. 1902). And I would continue to follow that rule today.

With this understanding of the relevant tax canons, I concur in the majority opinion.